Published : 50 seconds ago, on
By Kris Kowal
Origination, relationship management, underwriting, loan servicing, portfolio monitoring, risk management, compliance, innovation and product development: That’s commercial lending.
The question isn’t if, but rather when, generative AI touches every step of the commercial lending value chain. Commercial banks should be focusing on that “when.” That’s because, even if the GenAI tools were universally mature, secure, and reliable, AI talent limitless, and computing cycles gratis, no bank could manage a wholesale GenAI transition all at once.
The questions is, then, where to start—although most banks already have started. The question is definitely not “Why start?”
McKinsey estimates that GenAI productivity improvements alone could reap the equivalent of 9% to 15% of banking-industry operating profits. Accenture says it expects early-adopter institutions to enjoy 22% to 30% productivity improvements, a 6% rise in revenue growth, and 3% higher return on equity in the next three years.
Imagine running a race against competition that’s discovered a WADA-compliant supplement regimen that renders them 30% faster. That’s where we’re headed. These are, quite simply, staggering figures, and they jibe with bigger-picture analyses and survey results from the likes of BCG and Deloitte.
So, then, where does a commercial bank start/focus its energies during the opening laps of this long race?
GenAI projects are still IT projects
It depends. And that’s a good thing, because the work that goes into knowing where to start is familiar enough. GenAI implementation projects are IT projects, and banks know IT projects:
You get buy-in and support from the C-suite, because, as the above numbers show, these are strategic initiatives with existential competitive implications.
You define the problems GenAI might help solve and then prioritize them. Where are the most glaring deficiencies and opportunities? Are they on the processing side? The acquisition side? The servicing side?
You consider vendors. Are your current vendors embedding AI into their products? They probably are, so where are they focusing? This can offer clues as to where competitors and the industry at large recognize the greatest value, at least in the short-term, and that can help you prioritize. What would be involved in using new vendors—not only from the perspective of functionality, but also in terms of integration effort and data security?
You establish a data strategy. That involves determining what data you have, the quality of that data, where it resides, and how it might be accessed and harnessed. Also, if you’ve got straggling systems not yet in the cloud, now is the time to move them there, because GenAI is most powerful when it can access a wealth of real-time as well as historical data.
You establish metrics. What results are you targeting? Cost savings? Customer satisfaction? Turnaround time? Throughput? What are the KPIs to measure them?
And, finally, you determine how all this will affect workflows and, ultimately, the GenAI-assisted people in those workflows. That means working through the details of change management and upskilling.
A few examples of GenAI in commercial lending
As far as GenAI applications that commercial banks are focusing on, the prevailing approach has been to focus on efficiency and productivity first while keeping an eye on opportunities to innovate as the institution’s comfort with and skills in GenAI develop. To pick a just couple of examples:
Origination and underwriting. GenAI’s ability to tap into, parse, and synthesize diverse information sources plays well here. In addition to analyzing loan applications, GenAI can identify strengths or irregularities buried in years of financial statements, evaluate market trends related to the businesses applying for loans, and estimate the value of real estate and other collateral.
At the same time, GenAI can quickly assess how a prospective loan might affect the overall risk posture of a commercial bank and run scenarios on how that might change should the borrower default. These sorts of functions can be built into workflows or accessed via AI assistants responding to natural-language queries.
Loan servicing. GenAI’s power here can come from summarization—or from speedy, accurate rote presentation. When a borrower calls in with a question or issue, a relationship manager or customer-service agent can have the AI assistant summarize the borrower’s current and past loan information along with customer-service interactions to get right to the point. Conversely, if a question pertains to specifics on interest rate, prepayment period, penalty clauses, or other legally binding details, GenAI can pull up the contract language verbatim and point the client to the precise location in the documentation.
Product development. Here’s where GenAI’s potential as a commercial lending innovation engine manifests most obviously. New products could pertain to GenAI’s ability to suggest upselling and cross-selling opportunities. With access to fresh customer data plus historical borrowing patterns, seasonal trends, regional vagaries, and more, GenAI might predict that a customer might face a cash crunch in a couple of weeks and suggest a short-term loan to cover the gap. Alternatively, new products could involve entirely new loan categories that are especially relevant to particular classes of businesses with emerging needs at particular points in time – for, say, fleet owners converting to EVs and investing in onsite charging infrastructure.
GenAI is changing commercial lending for the better. It’s automating manual tasks. It’s providing immediate, reliable answers to a host of what used to be “I’ll have to get back to you on that” sorts of questions. And it’s fueling product innovation. That’s all just the start of what promises to be a rapid proliferation of GenAI across the full spectrum of commercial lending functions. Much of it will be new for banks. Heck, it’s new for everyone.
Fortunately, the IT development approaches that commercial banks have honed over decades also apply to GenAI systems. The key, as ever, is to understand your needs and your customers, find the right tools and partners, and solve the problems that matter to you most.
Kris Kowal is a senior industry expert in SAP’s global banking practice.